悠悠楠杉
网站页面
防红链接生成接口
03/10
首先，确保你安装了必要的库：
bash
pip install transformers
pip install nltk
```python
import nltk
from transformers import pipeline
from nltk.corpus import stopwords
nltk.download('stopwords')
stop_words = set(stopwords.words('english'))
初始化一个文本生成模型，这里使用GPT-2模型
summarizer = pipeline(
"summarization",
model="gpt2-medium",
topk=4,
minlength=30,
dosample=True,
stopsequence=[101] # 用于停止摘要生成的序列（对于GPT-2）
)
def generatearticle(title, keywords, description, content):
# 确保输入文本长度在1000字左右
if len(content) > 1000:
content = content[:1000] + " ... (继续阅读)" # 仅为了示例，实际中可以根据需要调整截断策略
# 生成基于关键词和描述的Markdown文章结构
articletext = f"## {title}\n\n### 关键词：{', '.join(keywords)}\n\n{description}\n\n---\n\n" + content + "\n\n"
return article_text
def generatesummary(text):
# 使用GPT-2模型生成摘要，并稍作调整以适应Markdown格式
summary = summarizer(text, minlength=50, maxlength=150, dosample=False)['summary_text']
return summary.replace(".", ".\n\n") # 添加换行以适应Markdown格式的阅读体验
示例数据输入（标题、关键词、描述、内容）
title = "The Importance of Digital Literacy in the 21st Century"
keywords = ["digital literacy", "technology", "education"]
description = "In this age of constant technological advancements, digital literacy has become increasingly important for individuals and society as a whole."
content = "Digital literacy is the ability to find, understand, evaluate, and create information in a digital format. It is not just about using a computer or a smartphone, but rather understanding the underlying concepts and principles of technology." # 保持长度在1000字左右（实际情况中，你需要更长的真实内容）
生成文章和摘要（可选）
fullarticle = generatearticle(title, keywords, description, content)
summary = generate_summary(content) # 这里仅生成摘要的示例代码，可根据需要选择是否使用或修改位置。
print(f"Generated Article (Markdown):\n") # 打印生成的Markdown文章（为了便于展示，这里使用多行字符串形式）
